Linux

On linux you can use gperftools

• install gperftools (available as a pre-built package on most platforms)
• make sure debugging symbols are being included in your build by, with g++ or clang++ adding -g to your CMAKE_CXX_FLAGS.

then if you are using a program that used the IMP flags support

• you can add the flag --cpu-profiling to profile the whole program

if not

• either use environment variables to control profiling as in the web page above
• or add -DCMAKE_EXE_LINKER_FLAGS=-ltcmalloc_and_profiler -DCMAKE_SHARED_LINKER_FLAGS=-ltcmalloc_and_profiler -DCMAKE_REQUIRED_FLAGS=-ltcmalloc_and_profiler to your cmake invocation to add the required library to all compilations. This will activate control via IMP flags and classes. Note that the name of the library might be different on your system. See the gperftool docs for more details.

To profile a specific piece of code use:

#include <IMP/benchmark/Profiler.h>

add a Profiler before the code you want to profile:

IMP::benchmark::Profiler pp("prof_out");

This will produce a prof_out file that can be viewed with pprof tool.

For detection of memory leaks, run your program this way:

env HEAPCHECK=normal your_program_exe

it will output the leaks and also produce output file that can be viewed with pprof.

Mac

On a Mac, you can use the Instruments program that is part of Xcode. It is pretty straightforward to use and can do CPU profiling, memory profiling, leak checking etc. It is currently located in /Applications/Xcode.app/Contents/Applications/Instruments.app/.